OSMscale (version 0.5.20)

earthDist: distance between lat-long coordinates

Description

Great-circle distance between points at lat-long coordinates. (The shortest distance over the earth's surface). The distance of all the entries is computed relative to the ith one.

Usage

earthDist(lat, long, data, r = 6371, i = 1L, along = FALSE, quiet = FALSE)

Value

Vector with distance(s) in km (or units of r, if r is changed)

Arguments

lat, long

Latitude (North/South) and longitude (East/West) coordinates in decimal degrees

data

Optional: data.frame with the columns lat and long

r

radius of the earth. Could be given in miles. DEFAULT: 6371 (km)

i

Integer: Index element against which all coordinate pairs are computed. DEFAULT: 1

along

Logical: Should distances be computed along vector of points? If TRUE, i is ignored. DEFAULT: FALSE

quiet

Logical: suppress non-df warning in getColumn? DEFAULT: FALSE

Author

Berry Boessenkool, , Aug 2016 + Jan 2017. Angle formula from Diercke Weltatlas 1996, Page 245

See Also

maxEarthDist, degree for pre-formatting, http://www.movable-type.co.uk/scripts/latlong.html

Examples

d <- read.table(header=TRUE, sep=",", text="
lat, long
52.514687,  13.350012   # Berlin
51.503162,  -0.131082   # London
35.685024, 139.753365") # Tokio
earthDist(lat, long, d)      # from Berlin to L and T: 928 and 8922 km
earthDist(lat, long, d, i=2) # from London to B and T: 928 and 9562 km

# slightly different with other formulas:
# install.packages("geosphere")
# geosphere::distHaversine(as.matrix(d[1,2:1]), as.matrix(d[2,2:1])) / 1000

# Distance along vector of points:
d <- data.frame(lat=21:50, long=1:30)
pointsMap(lat,long,d, zoom=2, proj=putm(1:30) )
along1 <- earthDist(lat,long,d, along=TRUE)
along2 <- c(0, sapply(2:nrow(d), function(i) earthDist(lat,long,data=d[i-1:0,])[2]))
along1-along2 # all zero, but second version is MUCH slower for large datasets


# compare with UTM distance
set.seed(42)
d <- data.frame(lat=runif(100, 47,54), long=runif(100, 6, 15))
d2 <- projectPoints(d$lat, d$long)
d_utm <- berryFunctions::distance(d2$x[-1],d2$y[-1], d2$x[1],d2$y[1])/1000
d_earth <- earthDist(lat,long, d)[-1]
plot(d_utm, d_earth) # distances in km
hist(d_utm-d_earth) # UTM distance slightly larger than earth distance
plot(d_earth, d_utm-d_earth) # correlates with distance
berryFunctions::colPoints(d2$x[-1], d2$y[-1], d_utm-d_earth, add=FALSE)
points(d2$x[1],d2$y[1], pch=3, cex=2, lwd=2)